Infant mortality under the age of one year had fallen from 8.4 million 1990 to 5.4 million 2010 according to the United Nations Children’s Fund (UNICEF). However, researchers from South Korea point out that this improvement has varied significantly by country depending on very different environmental factors infants face. Reducing the infant mortality rate is goal 4 of eight in the millennium development goals (MDGs) established by United Nations (UN). The team has attempted to predict the future infant mortality rate of 21 OECD countries from the year 2015 to 2050 using experience curve modeling and has compared the results to the two other well-known projections by the United Nations Population Division and the US Census Bureau in the context of the MDG targets. Their work suggests that as many as 18 of those 21 countries and no fewer than four of them may well not meet the targets by 2050.
Chang, Y.S., Lee, J. and Kwon, H.J. (2018) ‘When will the 2015 millennium development goal of infant mortality rate be finally realised? – Projections for 21 OECD countries through 2050‘, Int. J. Data Analysis Techniques and Strategies, Vol. 10, No. 1, pp.1-20.
The image compression formation known as JPEG (named for the Joint Photographic Experts Group) and often compressed to “jpg”, is a lossy compression format. A computer algorithm usually examines the pixels in an image, whether photograph or drawing, and makes decisions about discarding some pixels while retaining others and replacing repetitive information with a summary of the repeats in the image. Running such algorithms requires computer time and resources and so there is always a need to find ways to optimize the baseline algorithm especially as pixel numbers and detail in images rises with every iteration of camera and imaging technology. A team in India has applied a problem approximation surrogate model (PASM) to assist the differential evolution (DE) algorithm and found that they can reduce compression and decompression time of JPEG images relative to conventional approaches without losing any additional detail or quality in the process.
Vinoth Kumar, B. and Karpagam, G.R. (2018) ‘Reduction of computation time in differential evolution-based quantisation table optimisation for the JPEG baseline algorithm‘, Int. J. Computational Systems Engineering, Vol. 4, No. 1, pp.58-65.
Securing your Android
The Android operating system used by many mobile devices such as smartphones and tablets is growing in popularity but remains a target for malware writers, as does any popular operating system on any computer device. Researchers in China have turned to the cloud to help them protect Android devices. Their malware static detection scheme based on cloud security structure is both highly efficient and accurate in detecting malware that might otherwise compromise the operating system. Proof of principle was demonstrated against more than 1100 known malware apps with a pool of almost 3000 genuine apps. By using the cloud to carry out the scanning, the requisite processing power is kept off-device. An accuracy of 99.02% was observed with much lower demand on the device than conventional anti-malware software would require.
Yu, B., Song, P. and Xu, X. (2018) ‘An android malware static detection scheme based on cloud security structure‘, Int. J. Security and Networks, Vol. 13, No. 1, pp.51-57.
Social media solves crime
Could online social networks such as Facebook, LinkedIn, Google Plus and Twitter with their vast data sources be mined to detect and track crime? That’s the suggestion of researchers in the UK and their colleagues in Norway and Albania. Evidence might be gleaned from social media activity, for instance. The team has now developed an ontology, essentially a glossary, to facilitate forensic investigation of such evidence. The social media ontology, SMONT, gets around the problem facing investigators in that information from social networks has until now been mostly expressed in the form of classes rather than in terms of object properties. SMONT helps in the analysis of trends and behaviour of the public from the criminal investigator’s perspective.
Kalemi, E., Yildirim-Yayilgan, S., Domnori, E. and Elezaj, O. (2017) ‘SMONT: an ontology for crime solving through social media‘, Int. J. Metadata, Semantics and Ontologies, Vol. 12, Nos. 2/3, pp.71-81.